I see the discussion about the good and bad attributes of “Let Loose” Marshadow are having on the game and I wanted to just weigh in with that stuff I like: Math.
This will be a much briefer diatribe than my usual missive, but I wanted to put it out there for you because you deserve that good stuff.
The worst thing about Marshadow, I think everyone would agree, is a situation where the player who goes first has a strong turn, sets up well, and then plays a “Let Loose” Marshadow, setting both them and their opponent at 4 cards. The thinking in this situation is that the strong setup gives them a much higher likelihood of having a strong play coming out of the hand disruption than their opponent, who has not set up at all.
The key complaint that the opponent makes in a situation like this is that, by having their hand reset to 4 cards, they may, more or less, instantly lose by finding themselves with a dead hand. For the purposes of this discussion, we define a dead hand as a hand that has no way to draw additional cards and simply draws and passes back to their opponent. For a long, long discussion on this, let me refer you to my article on SixPrizes.
So here is what I did:
I calculated the Hypergeometric Distribution for various outs to draw using a 5-card (e.g. post “Let Loose” hand) and an 8-card hand. This is intended to simulate the situation where someone goes first, plays a supporter, then plays “Let Loose” and we see if their opponent has a dead hand.
This graph shows the odds of having at least 1 out to draw in your hand when you draw varying amounts of cards from a 59-card deck as you change the quantity of outs in the deck. As you can see, the odds of having a dead hand are much higher with a 5-card hand than an 8-card hand. As one adds outs to draw to your deck (e.g. more supporters), the delta shrinks, but it is still fairly substantial.
|Pop Size = 59||Cumulative Probability: P(X => 1)|
|Number of successes in population||Sample Size = 5||Sample Size = 8|
What is the significance of the gap between these two lines? Let’s call that gap “The Instant Loss Delta”.
This represents the difference in likelihood of a player having a Turn 1 out to draw going second if they have an 8-card hand versus a 5-card hand. Starting their game with a 5-card hand instead of 8, they have an 18% higher likelihood of instantly losing (e.g. drawing dead). If they have 18 outs to draw, they have a 10% higher likelihood of drawing dead and instantly losing.
Now, one could say “10% isn’t that bad!” But let me frame the problem in the way that players actually experience it: If you build a deck designed not to draw dead – e.g. many, many outs to draw – how does that shape your experience?
If someone does not play “Let Loose”, adding incremental outs to draw is frequently a poor investment. Just in the context of this conversation, adding an 18th out to draw improved your T1 draw engine by only 1%. Each incremental out to draw after 17 makes a difference in only (at most (on average)) 1 in 100 games. The alternative would be to replace that with a card that has more utility. I would posit that there is a card you could play that would impact a higher probability of games.
So countering “Let Loose” means compromising your deck in other matchups.
But it is actually much, much worse than that.
A slightly different way to look at this question would be: How do the odds change regarding the probability of drawing dead if you get “Let Loosed” T1? This is a different question, but illuminating regarding the challenge. If you only have 10 outs to draw, you draw dead with an 8-card hand more than 20% of the time. In other words, you expect to draw dead all the time. So when you get “Let Loosed” and draw dead 39% of the time, that is an 88% increase in the odds that you draw dead, but maybe you don’t take it so personally because you expected to draw dead anyway.
Now let’s imagine that you are terrified of getting “Let Loose” played on you, so you start increasing your outs to draw. If you have 17 outs to draw in your deck, you have an out to draw T1 94.7% of the time. You have to play 20 games before you draw dead. Alas, if you get “Let Loosed”, the odds that you draw dead are 17%, an increase of more than 220%! So if you are playing a deck with better insulation against drawing dead, the impact of Let Loose is RELATIVELY MORE SIGNIFICANT. In many respects, against a deck with many outs to draw, playing “Let Loose” is the best strategy an opponent can have because they go from having a 5% chance of instantly winning to a 15% chance! This means that there is no hard counter and the soft counters to this strategy are relatively ineffective.
The important thing to take away from this is that even the best-insulated decks draw dead 10% of the time off of a Let Loose, meaning that a single play creates a 10% chance of basically instant victory even against decks that are attempting to counter. What better play T1 could there possibly be?
Now let’s look at a few practical examples:
Isaiah William’s OCIC-winning Zapdos list has, in my mind, ~16 outs to draw:
- 4 Lillie
- 3 Cynthia
- 3 Professor Kukui
- 4 Nest Balls
- 1 Ultra Ball
- 1 Oranguru
This isn’t quite right because there are scenarios where you can’t discard enough cards to use Oranguru, you may have started Jirachi, or you might draw a Jirachi and have a switch effect, so maybe this is a conservative analysis, but let’s go with 16 for the sake of discussion.
If you “Let Loose” Isaiah T1, the odds that his hand bricks goes from 6.5% to almost 20%, a 197% increase! A “not particularly skillful play” gives you a 1 in 8-ish chance of winning immediately (the incremental growth in odds of dead draw). How could you not make that play?
Let’s look at Stephane’s 2nd place list:
- 4 Lillie
- 1 Cynthia
- 1 Judge
- 4 Ultra Ball
- 3 Pokemon Communication
- 2 Tapu Lele-GX
For this conversation, I didn’t count Mallow or Zoroark-GX in hand because that doesn’t change the productivity of your T1, it just ensures that your next turn is more productive. But I recognize that it helps. Regardless, where do you find yourself T1? You only have 15 outs to draw. The good news is that without your opponent playing “Let Loose”, you only draw dead 8% of the time. Unfortunately, if you get “Let Loosed”, you draw dead more than 21% of the time, a 171% increase. More than 1 in 5 games where you get “Let Loosed” end instantly, no matter how skillful you are. And adding a supporter only gives you an extra 2.4% of security in that situation. Hardly impactful.
Why doesn’t Judge have the same impact? Because it is your opponent’s supporter for turn.
The true imbalance of Marshadow only shows itself when your opponent exploits a set-up imbalance. If they play a bunch of cards, then Lillie for 8, then play a bunch of cards, and then Marshadow you, the odds that they draw dead are simply lower than your odds. Similarly, the odds that if they draw dead they will still be able to get attackers activated are much higher.
Here is some data on drawing 5 cards from a thinner deck:
|Number of successes in population||59 card deck||45 card deck|
As you can see, every card they burn that isn’t an out to draw improves their odds when they “Let Loose”. Similarly, even if they burn a few outs, assuming they conserve most of their outs (which is easy since “Let Loose” is shuffle/draw), the odds are still pretty strong! 13 outs in a 45 card deck is better than 17 outs in a 59 card deck.
Further, drawing dead next turn is simply less impactful to them. Imagine that they are playing a Malamar deck and they set up 3 Inkay and got an energy on an attacker. If they don’t draw a supporter but instead draw 2 Malamar, they are probably about to have a great turn. If they are playing a Zapdos deck and draw into a switch effect and Jirachi, they are going to have a great turn even though those are not “draw” supporters in the context of this discussion.
While I don’t know how to analyze this with data, I would also theorize that Marshadow has encouraged players to play Big Basic decks. This level of disruption early game impacts Stage 2 and other evolution decks relatively more than decks that attach and attack.
Marshadow decreases the benefit of skillful play by increasing the impact of RNG on the game. Marshadow decreases interaction in gameplay. Marshadow decreases the range of viable decks in the meta by decreasing the viability of evolution decks.
What is the benefit of Marshadow? It lets players push harder for resources and offers mild hand disruption late game in a format absent N.
I feel like one would be hard-pressed to make the case that the benefits of Marshadow outweigh the negative features it brings to the format.